Right ventricular function in human sepsis - A thermodilution study

Study objective: To assess the relations among right ventricular (RV) pressures and volumes in sepsis. Design: Prospective study. Setting: ICU. Patients: Eighteen patients with sepsis who were in hemodynamically stable condition. Intervention: Stepwise increases and decreases in RV end-diastolic volume (EDV) as induced by military antishock trousers (MAST) inflation of 0, 15, 30, 50 and then 0 mm Hg over 15-min intervals. Measurements and results: RV volumes and pressures were measured using a rapid response thermistor pulmonary arterial catheter based on RV ejection fraction (RVEF) calculations. RV EDV was estimated as stroke volume (SV)/RVEF, while KV end-systolic volume (ESV) was estimated as EDV-SV. Right atrial pressure (Pra) was taken as end-diastolic pressure, and pulmonary artery pressures (Ppa), both mean and end-systolic, were used as RV ejection pressures. MAST inflation to 15 mm Hg had no measurable effects on the measured variables. However, inflation to 30 and 50 mm Hg increased Pra, both mean and end-systolic Ppa, EDV, and ESV, whereas SV and RVEF remained. unchanged. The relation between either Pra and EDV (r=0.33) or the change in Pra and EDV, mean or end-systolic Ppa and ESV (r=0.5 and 0.28, respectively), car the change in mean or end-systolic Ppa and ESV for both the group and individual subjects was poor. Furthermore, SV correlated poorly with EDV (r=0.32), while RVEF correlated better with both EDV and ESV (r=-0.41 and -0.69, respectively), although it showed no definable relation to mean or end-systolic Ppa. However, both absolute and relative changes in EDV corresponded closely with respective ESV values (r=0.93). Conclusions: During sepsis, RV EDV and ESV vary independently of changes in Pra and ejection pressure. These data can be explained by assuming that the RV is a highly compliant chamber during filling, such that changes in RV EDV do not alter RV wall stress (preload) or ejection efficiency (RVEF). Thus, changes in RV EDV should proportionally alter RV ESV. Furthermore, the slope of the ESV/EDV relation should be inversely proportional to ejection efficiency.